FISHERY BULLETIN: VOL. 86, NO. 2 



Second pereopod (Figs. 5/i; 7a, b) shorter and 

 more slender than first, reaching to between mid- 

 length and end of antennal peduncle; fingers 

 slightly longer than palm, similar in size and 

 shape, opposed edges without gape, each pecti- 

 nate with single row of teeth in distal half di- 

 rected obliquely distad and increasing slightly in 

 size to end in noticeably stronger tooth crossing 

 opposite member when closed, but spineless prox- 

 imally; carpus slender, about 1.2 longer than 

 chela; merus and ischium unarmed. 



Third to fifth pereopods (Fig. 5i,j,k) similar in 

 length and structure, third reaching beyond an- 

 tennal scale by about 0.3 length of propodus. 

 Length articles of these legs in allotype 9 , mm: 



3rd 



4th 

 5th 



ischio- 

 merus 



4.48 

 4.89 

 4.16 



carpus propodus 



2.30 

 2.18 

 2.24 



3.20 

 3.39 

 4.22 



dactyl 



0.48 

 0.48 

 0.48 



Each short dactyl armed with about 6 corneous 

 spines on flexor surface, grading from small prox- 

 imally to longest and strongest distally; carpus of 

 each with distodorsal extension projecting as a 

 stop along proximal part of propodal extensor sur- 

 face; third leg stronger, at least in merus- 

 ischium, than fourth and fifth; merus of third and 

 fourth with ventral spine at 1/3 and 2/3 length, 

 distal one strongest, fifth without spines; ischium 

 of third with 2 spines in line with those on merus, 

 that of fourth and fifth spineless. 



Pleopods well developed, pair 1 with endopods 

 about half length of expods in both sexes, endopod 

 of male (Fig. 6j) with asymmetrical mesial exten- 

 sion, that of female tapering to acute tip; pair 2 

 with appendix masculina of male (Fig. 6^) bear- 

 ing distal cluster of about 9 strong straight spin- 

 ules extending beyond level of simple slender ap- 

 pendix interna. 



Uropod (Fig. 5e) with rami subequal in length, 

 slightly exceeding distal end of telson, lateral 

 ramus with movable spine mesial to smaller dis- 

 tolateral tooth, diaeresis sinuous. 



Etymology . — The name is from the Greek 

 "stactos", oozing out or trickling, and "philos", to 

 love, for association of the species with hydrocar- 

 bons seeping from the substrate. 



Remarks. — Alvinocaris lusca and the three 

 new species of Alvinocaris described here exhibit 



minor differences that are highlighted in the key 

 to species given above, but their similarities seem 

 far more significant; i.e., general body appear- 

 ance and strength of integument, shape of ros- 

 trum (although that of A. stactophila sometimes 

 lacks ventral teeth), shape and general armature 

 of tail fan, blindness, and general structure of 

 appendages, including mouthparts. Some minor 

 differences that may be mentioned are features 

 such as number of incisor teeth on the mandible, 

 number of spines on the first maxilla, shape of the 

 second maxilla, lack of meral spines on pereopod 

 5 in A. stactophila, unequal distribution of spines 

 on ischia of pereopods 3-5 in the three species, 

 and shape of the endopod of male pleopod 1 and 

 appendix masculina (though males of A. marken- 

 sis are not yet known). 



Each of these species lives in a distinctive ben- 

 thic environment, but all share similarities that 

 suggest dependence on a chemotrophic bacteria- 

 based food chain (Childress et al. 1986). Van 

 Dover et al. (in press) provide evidence from mor- 

 phology, behavioral and gut content analyses of 

 the similar Rimicaris exoculata Williams and 

 Rona that indicates a bacterial diet grazed from 

 surfaces of hydrothermal chimneys, although di- 

 rect observations of bacteria within the digestive 

 tract could not confirm this. The distinctively 

 spoon-shaped chelae of the first pereopods of both 

 Alvinocaris and Rimicaris species, with unbroken 

 comb of exceedingly fine teeth on the prehensile 

 edges, could be an adaptation for scooping or 

 sweeping bacteria toward the mouthparts. 

 Williams and Chace (1982) described the first 

 chelae of A. lusca as convex on the extensor sur- 

 face and concave on the flexor surface, but they 

 also said (p. 142) that the outer surface of the 

 fingers is convex and the inner surface concave. 

 The latter is misleading because in full extension 

 the convex side of the chela is mesial and the 

 concave side lateral. It is not yet known how these 

 appendages are used, but certainly the chelae can 

 be folded compactly against the leg's proximal 

 articles, and in the related Rimicaris exoculata 

 and R. chacei (Williams and Rona 1986) these 

 legs seem very mobile. Sensillae flanking prehen- 

 sile surfaces of the fingers seem well adapted to 

 aid feeding on finely particulate matter. More- 

 over, the species of Rimicaris have exceedingly 

 setose mouthparts. 



In species of both genera, the second pair of 

 pereopods have much smaller chelae with fingers 

 bearing long sensory setae and spines on the pre- 

 hensile edges that are seemingly adapted for 



278 



